Rhodium(II1) Complexes

Apr 13, 1977 - thank Dr. George Gray of Varian Associates for obtaining this spectrum. (9) The 3'P chemical shifts are downfield with respect to exter...
1 downloads 0 Views 560KB Size
3190 Inorganic Chemistry, Vola16, No. 12, 1977 ( 6 ) Other ylide complexes have been found to be useful for ethylene polymerization catalysts: US.Patents 2998416 (1961) and 3686 159 ( 1972). (7) All proton chemical shifts are reported with respect to Me4% (8) Due to the low solubility of 1 in any nonreactive solvent, a pulsed NMR technique was necessary to obtain this spectrum. The authors wish to thank Dr. George Gray of Varian Associates for obtaining this spectrum.

K. S. Bose and E. H. Abbott (9) The 3'Pchemical shifts are downfield with respect to external 85% H3PO4. (10) The details of these variable temperature IH NMR studies will be reported elsewhere. ( 1 1) Hydroformylations of ethylene, 1-hexene,and butadiene were attempted in benzene solution at 100 'C and pressures of CO/H2 ( 1 : I ) up to 1200 psi. At temperatures of 150 "C and above hydroformylation did occur, but at these temperatures the catalyst decomposed to give rhodium metal.

Contribution from the Department of Chemistry, Hunter College of the City University of New York, New York, New York 10021

Natural Abundance Nitrogen-15 Nuclear Magnetic Resonance Spectroscopy of Some Rhodium(II1) Complexes K. S. BOSE and E H. ABBOTT"

Received April 13, 1977

AIC70262Z

The natural abundance nitrogen-15 N M R spectra of the rhodium(II1) complexes of several alkyldiamine and aza aromatic ligands are reported. Spectra were obtained in several hours for the first class of compounds but required ca. 15 h for the latter group. Spectra could not be observed for a number of other complexes. The resonances of the complexes are found ca. 20 ppm upfield from those of the free ligand in the alkyldiamine case but are ca. 100 ppm above the aza aromatic ligands. This is attributed to shifts arising from the influence of low-lying excited states for the aza aromatic ligands themselves rather than any special property of their complexes When comparison is made between the chemical shift of the complex and that of the protonated ligand, a rather constant difference of ca. 30 ppm is observed. No correlation of shift with position of the ligands in the spectrochemical series is observed. I t is suggested that the upfield shift generally observed is a consequence of nonbonded magnetic interactions. Coupling constants between nitrogen- 15 and rhodium- 103 are observed and correlate simply with the s character of the bonding electron on nitrogen with J = 18 H z for sp2 and J = 14 H z foi sp3. Diastereoisomerism is detectable in the tris complex of 1,2-diaminopropane, and rather rigid, different ligand conformations are observed for the tris( 1,3-diaminopropane) case.

Introduction The nuclear magnetic resonance spectroscopy of nitrogen in coordination compounds is an area of considerable potential interest. Nitrogen is a constituent of many ligands which are important in coordination chemistry and biochemistry. Furthermore, unlike carbon and hydrogen, nitrogen is often bound directly to the metal ion, and its spectroscopy can be expected to be a sensitive probe of bonding, coordination geometry, electronic structure, etc. Nitrogen exists in nature as two isotopes. The most abundant one, 14N, has S = 1, and so its resonances are often quite broad in nonzero electric field gradients. In some situations this is not a problem, and its spectroscopy has yielded important information.' "N, however, has S = 'I2and so its resonances should be sharp and should reveal spin coupling when present. This isotope, unfortunately, is characterized by a rather low abundance (0.365%) and long relaxation times. These problems have largely restricted its natural abundance study to organic molecules which are liquids or are very soluble. A few studies have been performed with organometallic or coordination compounds, but these have been restricted to ligands enriched in 15N. For example, the chemical shifts of coordinated ammonia have been measured via an E N D O R method,2 and the mode of bonding was identified in a dinitrogen ~ o m p l e x . ~ The most thorough study of 15N shifts in coordination compounds has been the work carried out using 15N-enriched ethylenediaminetetraacetate (EDTA) and a series of closed-shell metal ions. In this study it was demonstrated that shifts of only a few parts per million are observed when EDTA binds such closed-shell metal ions as Ca2+,Zn2+, Mg2+, etc. This disappointingly small shift was attributed to near cancellation of large local diamagnetic and paramagnetic effects a t the nitrogen atom. * To whom correspondence should be addressed at the Department of Chemistry, Montana State University, Bozeman, Mont. 59715.

Herein we report the natural abundance nitrogen-15 spectroscopy of a series of rhodium(II1) complexes. This metal ion was chosen for a variety of reasons. Its complexes are always diamagnetic and substitutionally inert, excluding ground-state paramagnetic or ligand-exchange effects from consideration. As a d6 ion, however, it has fairly low-lying excited states whose admixture might lead to large chemical shifts and whose energies are readily varied by utilizing ligands from different parts of the spectrochemical series. Also, it exists entirely as Io3Rh with S = 'I2,and so one can obtain coupling constant data. Experimental Section Abbreviations of ligands appear in footnote 5. The rhodium(II1) compounds Rh(en),Cl;, Rh(pn);C13, and Rh(tn);C13 were prepared by the method of Gatsbo16 while Rh(bpy)3Cl3 and R h ( ~ h e n C13 ) ~ were prepared by the method of McKenzie and Plowman.' Rh(en)2C12C1 and Rh(py),C12C1 were prepared by the methods of Anderson and Basolo' and of Gillard and Wilkinson,' respectively. The "N N M R spectra were obtained in Fourier Transform mode with JEOL PS/PFT-100 spectrometer interfaced to a JEOL EC-100 data system. Chemical shifts at 10.09 MHz were determined using a 5 kHz range and 8K memory and are measured with respect to an external ammonium chloride solution. In the case of aliphatic amine complexes, a pulse angle of about 60' and in the case of aza aromatic complexes, a pulse angle of about 25-30' was used. For all complexes H 2 0was the solvent, except in the case of the pyridine complex which was dissolved in benzyl alcohol. The spectra of the bis(ethy1enediamine) and tetrakis(pyridine) complexes were obtained at 50 'C to have sufficient concentration.

Results and Discussion Table I presents the chemical shift data obtained from the present studies. It also indicates the experimental conditions necessary to obtain these spectra. The most favorable cases are those of the bidentate saturated amines, which have protonated nitrogen atoms. Spectra are easily obtained within 2-4 h a t concentrations of 1-2 M. Figure 1 is the spectrum

lSNNMR

Inorganic Chemistry, Vol. 16, No. 12, 1977 3191

Spectra of Rh(II1) Complexes

Table I. Nitrogen-15 Chemical Shifts of Ligands and Their Rhodium(II1) Complexes and Rh-N Coupling Constants' Complex

Protonated ligand, P P ~

Rh(en),Cl, Rh(pn) ,C13 Rh(tn) $1, R~(~PY),C~, Rh(phen) $1, (trans- Rh(en) ,Cl,)Cl (trans-Rh(py),Cl,)Cl (cis-Rh(bp y ),Cl,)Cl (cis-Rh(phen) ,Cl,)Cl R ~ ( P ),(SCN), Y Rh(CH,CN),Cl,

8.6 21.7 7.7 10.3 195.4 211.4 8.6 176.9

-

Free ligand, PPm

Coordinated ligand, ppm

-6.5' 9.8' -9.4 -1.7' 280.7' 273.1d -6.5' 289.6' No spectrum observed No spectrum observed No spectrum observed No spectrum observed

-31.8 - 15.5 -34.5 -44 186.1 180.8 -28.6 191e

JRh-N,

HZ

Scansf

14.6 (13 .5)g

800 29 000

(14.3)g 18.1 18.4 13.4 17.1

30 000 10 000 25 000 5 000 46 000 50 000 50 000 50 000 5 0 000

a Shifts are downfield from an internal capillary of lSNH,C1. In CDCl,. In CH,OD. e At 50 "C. f Number of scans In H,O-HC1. to obtain a t least 10:1 signal to noise for 2 M or saturated solution. g Average coupling constants f o r t h e isomers present. 100 Hz

Table 11. Chemical Shift on Replacement of a Proton by Rhodiurn(II1) Protonated ligand enkFz2+ pnH,% tnH;+ phenH+ bpyH+ PYH+

Figure 1. I5Nspectrum of Rh(en)$13 at 10.09 MHz: 1.5 M solution in H,O; 800 scans with a pulse angle of ca. 60'.

obtained for the tris(ethy1enediamine) complex. The aromatic heterocyclic amines, however, lack directly bonded protons, and obtaining the spectra of their complexes is significantly more difficult. Here spectra could only be obtained for the most symmetrical complexes and then only with accumulations of 10 000 or more scans. Table I also lists a number of cases where no spectra could be obtained. All these complexes lack protons on the nitrogens and their symmetries are low. The relatively short times required to obtain spectra of complexes whose nitrogen atoms have directly bonded hydrogen is probably the result of signal enhancement due to the nuclear Overhauser effect resulting from proton noise decoupling plus the shorter relaxation times enjoyed by such nitrogen atoms. Paramagnetic relaxation reagents have been used to shorten relaxation times in carbon-1 3 spectroscopy; however, the addition of tris(ethylenediamine)chromium(III) chloride did not help in cases where no spectra had been obtained. Chemical Shifts. In Table I, a comparison is made among the chemical shifts of the complexes, the free ligands, and the protonated ligands. Chemical shifts are interpreted by means of the Ramsey equation, which treats the observed shielding as a combination of three terms as in eq l I o . The up term is the paramagnetic component and consists of a summation of terms whose numerators are matrix elements between ground and excited states and whose denominator is the difference in energy between the ground and excited state. It is a number of negative sign and results in shifts in the downfield direction. The ud term is the diamagnetic component arising from the electron density around the atom in

S , P P ~

8.6 21.7 1.7 10.3 211.9 195.4 176.9

Complex Rh(en),Cl, (Rh(en),Cl,)Cl Wpn),Cl,

8 , ppm

-31.8 -28.6 -15.5 -34.5 Rh( tn) ,C1, -44 Rh(phen) ,C1, 180.8 186.6 Rh(bPY),Cl, ( R ~ ( P Y ) , C ~ , ) C ~ 191.0

Diff, PPm -40.4 -37.2 -37.2 -42.2 -54.3 -30.6 -9 + 14.1

question in its ground state and is in the upfield direction. The ul term is a term arising from the effects of neighboring atoms. This term is often ignored; however, in cases where the nucleus in question is next to a transition metal, the term can be rather large. In the transition-metal hydrides, upfield shifts of 20-30 ppm are observed for the hydride and are the result of the ~1 term.' In Table I, an upfield shift is observed when the free ligands coordinate rhodium(II1). In the cases of the saturated ligands, the shift is 25 to 42 ppm; however, the shift is approximately 100 ppm for the aromatic ligands. Table I also contains the chemical shifts of the protonated ligands. Inspection shows that the different behaviors of aliphatic and aromatic ligands on coordination are similar to their different behaviors upon protonation. A brief digression on the effects of protonation is therefore instructive. Table I shows that for the saturated ligands there is a 10-15 ppm downfield shift when the free base is protonated, whereas shifts as much as 100 ppm upfield are noted when the aromatic ligands are protonated. These observations are consistent with results obtained for I5N shifts in analogous systems. For aliphatic amines, the downfield shift upon protonation has been attributed to changes in the a d term of eq 1. Thus, protonation is thought to cause a downfield shift mainly by withdrawing electron density and decreasing the diamagnetic component of shielding.12 For aromatic amines, the free ligand is found to lower field than the protonated ligand. Here, it is believed that the paramagnetic term is primarily responsible. The excitation energy for the lone pair of electrons on the nitrogen of the free base is much lower than when this pair binds a proton. Hence a much larger paramagnetic term is anticipated for the free base than for the protonated ligand.13 Thus, the difference between the behavior of aromatic and aliphatic ligands upon protonation has been traced to the low excitation energy of the aromatic free base lone pair. Lone-pair effects in the comparison of ligands to complexes may be eliminated from consideration by instead compairing protonated ligands to the metal complex. Table I1 makes such a comparison. It shows that rather similar upfield shifts are

'

3192 Inorganic Chemistry, Vol. 16, No. 12, 1977 observed for both aromatic and aliphatic ligands when rhodium(II1) replaces a proton. Roberts and co-workers have studied the 15N resonances of a large group of closed-shell metal complexes of EDTA.4 Shifts of less than 20 ppm were observed for all complexes. Utilizing an approximate formula suggested for the calculation of diamagnetic shielding in small molecules, these authors assumed neglible polarity of the metal-ligand bond and calculated values of ffd which are large and increase sharply with the atomic number of the metal. To rationalize the relatively small shifts observed, they postulated that a large paramagnetic term must also arise from coordination and that its magnitude must nearly equal the diamagnetic term, almost canceling it. Since the metal ion is the same in all the cases we report, its contribution to the diamagnetic term should be fairly constant from complex to complex. The metal ion's contribution to upcould be large because it has an unfilled shell and low-lying excited states. If such a upterm were large, one would expect to observe chemical shifts which are strongly dependent on the type of ligand because the energies of the excited states are determined by the position of the ligands in the spectrochemical series. Since the up term has the excitation energy in its denominator, one would expect an inverse relationship between up and the average position of the ligands in the spectrochemical series. If for a series of complexes up is large and g d is constant, the observed shielding should correlate with the spectrochemical series.I4 N o such correlation can be made from the data in Table I. The tris( 1,3-diaminopropane) (tn) complex, for example, is found to higher field than the en complex. However, for the corresponding cobalt(II1) complexes, tn is lower in the spectrochemical series and its complex has the longer wavelength absorption. Also, trans-dichlorobis(ethy1enediamine)rhodium(II1) falls only 3 ppm to lower field than the tris(ethy1enediamine) complex, despite the much lower position of chloride in the spectrochemical series. It appears that the up term for the effect of the metal ion on the ligand nitrogen atom is not large in the complexes we have studied. Since the shifts observed on coordination are small, ud must also be in the range of 10-30 ppm, as opposed to the values greater than 100 ppm suggested by Roberts. The smaller values are more in keeping with the 10-ppm shift observed when saturated amines are protonated. T h e changes in up and ud a t nitrogen are evidently small when a proton is replaced by a rhodium(II1) ion. It remains to suggest a plausible explanation for the relatively large and constant upfield shift observed upon coordination in Table 11. A simple explanation would lie in ul of eq 1. It seems likely that these upfield shifts are similar to those observed in proton spectroscopy when a transition-metal hydride is formed. Here it has been shown that the d orbitals of the metal ion interact with the magnetic field to shield the attached proton." Thus, the model we suggest is one where the nitrogen nucleus is not affected much by coordination to a metal ion except that it finds itself adjacent to a magnetically active neighbor which makes a rather large contribution to ul. Long-range shielding of this type depends inversely on the excitation energy. It also depends inversely on the cube of the internuclear separation. The effect should parallel the spectrochemical series if the metal-nitrogen bond lengths were constant. The complexes we have studied differ substantially from one another. Bond lengths vary considerably, and this is magnified when they are cubed. Therefore, it is unwise to make a more quantitative comparison among the complexes in order to evaluate u,. This will require the study of a much more closely related series of compounds. Coupling Constants. The coupling constants are easily rationalized in terms of the metal-nitrogen bonding. Table

K. S . Bose and E. H. Abbott 100 Hz

.c----+ n

Figure 2. 15Nspectrum of Rh(tn)jCl, at 10.09 MHz: 2 M solution in H20; 30000 scans with a pulse angle of ca. 60'.

I shows that saturated nitrogens couple to lo3Rh with J = 13-14 Hz, while the aromatic nitrogens couple with J = 18 Hz. The ratio of these values is nearly exactly that of the ratio of s character in pure sp3 and sp2 hybrids. Equation 2 is the

theoretical expression for nuclear spin coupling, where AE is the singlet-triplet excitation energy, the a's are the s character of the hybrids a t each atom, and the rc/(O)'s are the electron densities a t the respective nuclei. It is noteworthy that this metal ion excitation energy does not vary enough tc influence the simple ratio of coupling constant to hybridization. Similar observations have been made in studies of the coupling between platinum- 195 and directly bonded nuclei. In platinum phosphines, the magnitude of JIPSpt-Slp is independent of excitation energy and determined mainly by the bonding of the complex.I6 Likewise, 13C-195Pt couplings are governed by hybridization of the carbon atom" and 15N-195Ptcouplings are dependent on the ratio of s character between the hybridization of the platinum in octahedral and square-planar complexes. * Chemical Effects. Spectra were usually unobservable under our conditions for complexes with nonequivalent nitrogen atoms. The saturated amines were the only exceptions. Ethylenediamine chelate rings are nonplanar, leading to nonequivalence of the nitrogen donors; however, separate resonances aye not observed, no doubt because of the rapid intraconversion of isomers. For 1,2-propylenediamine complexes, diastereoisomers are expected for the tris complex. These were observed, but the chemical shifts between isomers are small, on the order of only 2-3 Hz. For 1,3-propylenediamine, the coordinated ligand may have boat, chair, and skew-boat conformations. The tris cobalt(II1) complex has equivalent ligands in the solid state,I9 though an equilibrium may exist with skew-boat forms in solution.20 The somewhat larger chromium(II1) has two equivalent tn ligands in chair forms facing one another and a third one in skew-boat conformation.21 Equilibrium among the ligands is expected to be rapid. Figure 2 shows the 15Nspectrum of the (tn)-Rh"' complex. Though the structure of the complex cannot be inferred, it is clear not only that the ligands are nonequivalent but also that they intraconvert slowly on the N M R time scale. The carbon-13 spectrum shows the same multiplicity. By contrast, the 13C spectrum of the corresponding cobalt(II1) complex is simply two singlets, one for each of the different carbon atoms. Conc1usions Natural abundance 15N spectra may be obtained for rhodium(II1) complexes if the nitrogens have directly bonded

Metalation of Dimethylaminomethylferrocenes protons. If the ligands do not have directly bonded protons, spectra are obtainable for large ligands in relatively symmetrical environments. The chemical shifts are not large and are usually to high field. No simple rationalization of the shifts can be made; however, the constancy of values observed suggests that they are largely due to nonbonded effects. Coupling constants appear to be sharply dependent on the hybridization of the nitrogen and may afford a convenient measure of the s character of metal-nitrogen bonds.

Inorganic Chemistry, Vol. 16, No. 12, 1977 3193 (3) J. E. Bercaw, E. Rosenberg and J. D. Roberts, J . Am. Chem. SOC., 96, 617 11974)

(4) R.%agen,

>. P. Warren, D. H. Hunter, and J. D. Roberts, J . Am. Chem.

SOC.,95, 5712 (1973).

(5) The abreviations used are: en = ethylenediamine; pn = 1,2-diaminopropane; tn = 1,3-diaminopropane; bpy = 2,2-bipyridine; phen = 1,lO-phenanthroline; py = pyridine. (6) F. Gatsbol, Inorg. Synrh., 12, 269 (1970). (7) E. D. McKenzie and R. A. Plowman, J . Inorg. Nucl. Chem., 32, 199 (1970). (8) S. N. Anderson and F. Basolo, Inorg. Synth., 7, 214 (1963). (9) R. D. Gillard and G. Wilkinson, Inorg. Synth., 10, 64 (1967). (10) N. F. Ramsey, Phys. Rev., 91, 903 (1953). Acknowledgment. This research was supported by Grant (11) A. D. Buckingham and P. J. Stephens, J . Chem. SOC.,2747 (1964). (12) (a) J. B. Lambert, G. Binsch, and J D. Roberts, Proc. Nutl. Acud. Sci. AM15707 from the National Institutes of Health and Grant U.S.A.,51, 735 (1964); (b) W. M. Lichtman, M. Alei, Jr., and A. E. GP-37025 from the National Science Foundation. Florin, J . Am. Chem. Soc., 82, 6574 (1960). (13) D. Herbison-Evans and R. E. Richards, Mol. Phys., 8, 19 (1964). Registry No. Rh(en)3C13, 14023-02-0; R h ( ~ n ) ~ C14175-72-5; l~, (14) R. Freeman, G. R. Murray, and R. E. Richards, Proc. R. Soc. London, Rh(tn)3C13, 64175-36-6; Rh(bpy),C13, 32680-72-1; R h ( ~ h e n ) ~ C l ~ , Ser. A , 242, 455 (1957). 5 ) R. Bramely, B. N. Figgis, and R. $. Nyholm, J. Chem. SOC.A, 861 (1967). (1 27353-21 -5;(trans-Rh(en),Cl,)Cl, 15444-63-0; (tr~nr-Rh(py)~Cl~)Cl, (16) A. Pidcock, R. E. Richards, and L. M. Venanzi, J . Chem. Soc. A , 1707 14077-30-6; enH2*+, 22534-20-9; pnH2’+, 62063-19-8; tnHzZ+, (1966). 61696-59-1; phenH+, 22559-75-7; bpyH+, 20755-72-0; pyH+, (17) M. H. Chisholm, H. C. Clark, L. E. Manzer, and J. B. Stothers, Chem. 16969-45-2; en, 107-15-3; pn, 78-90-0; tn, 109-76-2; bpy, 366-18-7; Commun., 1627 (1971). phen, 66-71-7; py. 110-86-1; 15N, 14390-96-6. (18) P. S. Pregosin, H. Omura, and L. M. Venanzi, J . Am. Chem. Soc., 95, 2047 (1973). (19) T. Noumura, F. Marumo, and Y. Saito, Bull. Chem. Sot. JDn., 42, 1016 References and Notes (1969). (1) N. Logan in “Nitrogen N M R , M. Witanowski and G. A. Webb, Ed., (20) P. G. Beddoe, M. J. Harding, S. F. Mason, and B. J. Peart, Chem. Commun., 1283 (1971). Plenum Press, New York, N.Y., 1973. (2) J. W. Lehman and B. M. Fung, Inorg. Chem., 11, 214 (1972). (21) F. A. Jurnak and K. N. Raymond, Inorg. Chem., 13, 2387 (1974).

Contribution No. 3732 from the Department of Chemistry, University of California, Los Angeles, California 90024

Metalation of Acetyl- and Dimethylaminomethylferrocenes with Pentacarbonyl(methy1)manganese or -rhenium: Formation of Homoannular Metalated Ferrocenes and Ferrocenylmethylaminomethylene(tetracarbonyl)manganese1~2 S A L L Y S C H R E I B E R C R A W F O R D and H E R B E R T D. KAESZ* Received November 6. 1977

AIC60804X

The homoannular metalation products tetracarbonyl(2-acetylferroceny1)manganese (3) and -rhenium (2) and tetracarbonyl(2-dimethylaminomethylferrocenyl)rhenium(9) have been isolated and characterized. The cyclometalated product 3 is found to be much more labile to reaction with donor molecules than the phenyl analogue, tetracarbonyl(2-ac$Apheny1)manganese. While 3 readily reacts with C O to form pentacarbonyl(2-acetylferroceny1)manganese (4) the phenyl compound fails to show any reaction even a t elevated temperature. While both species react with triphenylphosphine to form the phosphine-substituted facial tricarbonyl (e.g., fac-tricarbonyl(triphenylphosphino)-2-acetylferrocenylmanganese), 5, the phosphine substitution occurs more rapidly for the ferqocenyl analogue. Treatment of benzoylferrocene with pentacarbonyl(methy1)manganese leads solely to metalation on the phenyl ring giving tetracarbonyl(2-ferrocenylcarbonylpheny1)manganese (7). The cyclometalated ferrocenylamine 9 undergoes electron oxidation with FeC1,. Cyclic voltammetry indicates a reversible oxidation at +0.17 V (vs. SCE), lowered from that in the free amine (+0.43 V vs. SCE). The reaction of pentacarbonyl(methy1)manganese with dimethylaminomethylferrocene leads to a cyclometalation product involving the methyl group on nitrogen, Le., tetracarbonyl(N-ferrocenylmethyl-N-methylaminomethy1ene)manganese (10).

Introduction Up to the present study, the great majority of aromatic metalation reactions by transition metals have been observed for functionalized phenyl ringse3 The present work was initiated to explore the metalation of substituted metallocenes, of which only a limited number of examples can be cited. It was earlier reported that FcCH=NPh [Fc = ferrocenyl, ($-C5H5)Fe($-CsH4-)] failed to undergo metalation by Fe2(C0)9analogous to that observed for PhCH=NPh (eq 1):

As suggested by these authors, the negative result for the metallocene is most likely due to the unavailability of the cyclopentadienyl ring T electrons for interaction with the

second iron tricarbonyl moiety, a feature required in metalation by Fe2(C0)9of the Schiff base. That metallocene or metal-r-bonded arene rings can be u metalated by transition-metal complexes has since been demonstrated by report of a variety of complexes, I-V,5,6 while SeiwelPf has postulated metalated intermediates in the catalyzed H / D exchange in ferrocene and in cyclopentadienylrhodium compounds. During the study described here and independently of it, A l ~ e has r ~ reported ~ the metalation of thiopivaloylferrocene with Na2PdC14giving product VI. More recently, Gaunt and S h a have ~ ~ observed ~ cyclometalation of dimethylaminomethylferrocene by PdC142-. It should be noted as well that lithiation, and to a lesser extent sodiation, of substituted ferrocenes has been extensively studied.8 Through such metalated derivatives, transition-metal substituents may be brought in via the usual metathetical reactions; the mono- and bis[tris(~5-cyclopentadienyl)uranium] ferrocenesgmay be cited as examples of syntheses using such a pathway.